Two return pumps - Parallel or Series?

[LobsterOfJustice]

Would like some input on the best way to run two return pumps. My reasons for running two return pumps are:

1. Additional flow rate beyond the ability of a single pump.
2. Redundancy in the case of failure.

I understand the ideal situation is to run two pumps with completely independent plumbing all the way from sump to tank. This is not possible in my situation - I have a remote sump with plumbing run through the crawlspace by a professional plumber - I have a single 1.5" return line between the garage and the display on the other side of the house to work with.

As I figure it, my options are as follows:

1. Run two pumps in parallel. I am leaning toward this option. Each pump will have a check valve immediately after the pump, and will run 1.5" plumbing on the output before combining both lines (with a wye, not a tee) into a 2" line. I will manifold off this line to feed aux equipment and run it through a UV sterilizer before reducing back to 1.5" for the long run back to the tank. The installation of a check valve after each pump should mean that in the event one pump stops working, the other will continue returning water to the tank (and not back to the sump through the other pump's plumbing).

2. Run the pumps in series. The outlet of one pump would be 1.5" plumbed into the inlet of the second pump. Then I would run 2" plumbing out of the second pump and the remainder of the system would be same as above (manifold off line, UV, then reduce to 1.5" for run back to tank).

I've been going over this in my head and I'm not sure if running the pumps in series would give me the added benefit of additional flow rate. However, I'm also a little concerned that combining the outlets of the pumps into a single line (even though I will be up-sizing pipe at this point) may also not work out as well as I want (pumps may end up fighting against each other instead of working together). Running the pumps in series has the benefit of not requiring a check valve on each pump in order to provide backup flow in the case of single pump failure.

Anyway, I would love to get some input from people with experience with either of these plumbing configurations. Often times I think the fluid dynamics will work one way and then I get it all plumbed up and something unexpected happens... would prefer to get this right the first time.

 

[Maacc]

Definitely parallel, the only way series might help is if you had an enormously long run/high head pressure and such a system would be very vulnerable to failure if the first pump failed and did not get enough water to the second pump causing it to fail.
In your situation a lock up of either pump could cause a failure in the other either by cutting off flow or causing huge head pressure depending upon which pump failed and how.

 

[mcarroll]

Parallel
If you can't plumb them simultaneously, then parallel isn't even an option. This is the main setup you'd want for redundancy.

Serial
I wouldn't bother with serial as that's mostly good for increasing pressure.

The only exceptional case I can imagine serial plumbing being worth your while is if you got two aweome small pumps for almost-free and you needed to stack them for the added performance just to make them work for you.

Even in that case I think it would probably be better to sell them and get a right-sized pump. If backup is important, I'd still want a second one on-site....ideally plumbed in parallel.

Backup
As long as you have a second return pump on hand so you can replace yours when it fails without a protracted wait, you're already golden. (Return pumps only fail when your LFS is closed.)

Use Case Matters
Your use case matters. While you did start off mentioning an ideal, I can't tell what your use case is, so I might make different suggestions if I knew the goal of the plan.

I will say that so far I've never seen a tank where having multiple return pumps at once was really called for.

I have seen it done a few times, but those were not typical installations at all....one was someone trying to make "a return pump" that would work from the basement out of a pile of used MaxiJets. Another was just pulling a stunt by using two expensive pumps that weren't actually appropriate for the installation. And the last was upgrading an old system and had extenuating circumstances...neither redundancy nor increased GPH were goals.

2¢

 

[LobsterOfJustice]

In your situation a lock up of either pump could cause a failure in the other either by cutting off flow or causing huge head pressure depending upon which pump failed and how.

Are you talking about the series or parallel setup here?

Parallel
If you can't plumb them simultaneously, then parallel isn't even an option. This is the main setup you'd want for redundancy.

I don't exactly understand what you are saying here. While I cant plumb the two pumps back to the tank completely independently, I can plumb their inlets/outlets separate and then combine the outlets before running back to the tank.

Backup
As long as you have a second return pump on hand so you can replace yours when it fails without a protracted wait, you're already golden. (Return pumps only fail when your LFS is closed.)

I'll one-up your "Murphy's law" observation and say return pumps also tend to fail when you're out of town. I am wanting an immediate backup without physical replacement of pump. I am going out of the country for a little over a week in the fall and if one of the pumps fails, the second can keep the tank going on reduced turnover until I return. In this scenario, a replacement pump does me no good sitting on a shelf in the garage.

Use Case Matters
Your use case matters. While you did start off mentioning an ideal, I can't tell what your use case is, so I might make different suggestions if I knew the goal of the plan.

Sorry, I thought I was pretty specific as far as the goals of my setup. What other information would be helpful?

I'm looking for the ability to achieve 1000+ GPH after head losses (~8ft vertical, ~70ft horizontal) and after teeing off to feed media reactors, chiller, frag tank, denitrator, and refugium. I want the pump to be DC as I like to vary the output of the return pump to create a surge tank effect (not looking for comments on this strategy - been doing this for a few years now and extremely happy with the results). The problem is, once you look for a DC pump with a high enough GPH rating for my situation, there is a tremendous jump in price. And purchasing a single pump doesn't solve the redundancy concern. Specifically, right now I have a Waveline DC12000. I'm only using this to feed the media reactors and refugium so far. I'm decently satisfied with the amount of actual flow I'm getting back to the tank but would like the ability to have more, and at this point if I manifold in the additional items I will have even less flow making it back to the tank. I am looking to add a Varios-6 to make up for the flow lost due to manifolding in a few extra items and to give me a little extra oomph returning to the tank as well.

 

[mcarroll]

I am going out of the country for a little over a week in the fall and if one of the pumps fails, the second can keep the tank going on reduced turnover until I return.

I'm old-fashioned on this – you need a tank-sitter if there's a real concern.

Either that or you need a lower stocking level in the display that would be fine if the return failed for a week.

But you probably need a tank-sitter.

Unless you know a good reefer who's a friend, I'd consider hiring a pro....interview folks until you find someone you like.

If the tank's "not worth that" then just use a reliable return pump (use quality gear that has a track record, generally speaking) and try not to worry so much.

I don't know your local scene, but most folks have no issues like you're worrying about, BTW.

IMO...

Worry about power outages while you're gone....that's about it. Without a tank-sitter to be your hands and eyes while you're gone, however, there's only so much that can be done before-hand...an extended power outage is going to defeat any simple backup plan.

 

[Maacc]

I was referring to the series application.
You could do parallel and just leave it off on an apex to turn on while you travel if the first failed.

 

[LobsterOfJustice]

I will say that so far I've never seen a tank where having multiple return pumps at once was really called for.

I have seen it done a few times, but those were not typical installations at all....

Well, now you can't say that anymore

Following up, I plumbed in my second return pump in parallel and I am happy with the results. I used programming to adjust the pumps to roughly the same output of ~450 gph each. When both pumps are running at this programming the total flow is ~635 gph. That means there's about a 30% efficiency loss in terms of GPH but it's worth it to me for the redundancy and control. Using the apex I periodically ramp both pumps to 100% for a surge action and I am getting just shy of 1000 gph. Was hoping for more here but didn't feel like spending the extra for the Varios-8 over the 6.

If a return pump fails, the tank will still receive ~2/3 of it's normal flow from the other pump. If I am unable to deal with the issue in person within a few hours, I can adjust the programming of the remaining pump to make up the difference as well (since both are running at less than full power). As a bonus, each return pump is plugged into a separate GFCI circuit - so if one circuit trips, the other pump is unaffected.

 

[clsanchez77]

Two pumps in parallel is doable - I have done this on past tanks.

A few scenarios come to mind where you would want two pumps but you cannot accommodate separate return lines, or maybe you run a manifold and you don't want two of everything on that manifold manifold. I did this on previous tanks without issue. It even saved me on three occasions in the past 15 or so years. Here is a few notes.

• Must have check valves (Oversize them, check valves have a K value of 2.5!)
• Piping must be adequately sized otherwise the pumps will just dead head each other and you will have no net increase in flow. Size the piping for both pumps running and try to keep the velocity inside the pipe around 3-ft/s. We work with higher velocities in industry but the pumps we have in the hobby don't respond well to even slight increases in head loss.
• Pumps must have a reasonable pump curve, otherwise they will just dead head each other and you will have no net increase in flow. Pumps with steep drop offs wont work for this application unless you keep the head losses very minimal.

In almost every case, two pumps on separate returns will be much better. But I can certainly appreciate that cannot always be doable. I wanted to run parallel pumps on my current tank and I'm using a 1-1/2" manifold. No matter which way I turned, stacked, crammed, etc I just could not get the second pump in there. But if I could, I certainly was not going to get two manifolds in the stand, my tank is only 18" front to back. Luckily I run an Iwaki so in theory it should never fail

I cannot see any reason to run pumps in series. Unless you have your gear in the basement and the DT up top and the pump you really want to have just doesnt have the head. I personally think you should pick a better pump for the job and there is plenty.

And btw, Lobster, you have been around a very long time. I enjoyed following you on RC back in the day Glad to see you are still around!

 

[Grey Guy]

I plumbed 2 iwaki's paralell, into the same discharge and put a check valve on each pump discharge and plumbrd both pumps into the same intake. This way I could run 1 or both pumps without back feeding through the intake line. Would send a picture, but that system is no longer set up.

 

[ReeferBob]

I don't see the reverse flow preventer you mentioned above. If one pump stops, the other I think would send most of its flow to the other pump. Could cause a flood. Did you test that scenario?

 

[LobsterOfJustice]

I don't see the reverse flow preventer you mentioned above. If one pump stops, the other I think would send most of its flow to the other pump. Could cause a flood. Did you test that scenario?

I have a check valve installed directly on the output of each pump. You can see if you look closely at my second picture.

 

[Grey Guy]

I don't see the reverse flow preventer you mentioned above. If one pump stops, the other I think would send most of its flow to the other pump. Could cause a flood. Did you test that scenario?

Water will flow backwards through a pump pretty well and even make the impeller spin the wrong way.

 

[clsanchez77]

The check valves are clearly visible in his picture above.